Ambient vibration analysis on seismic arrays to investigate the properties of the upper crust: an example from Herdern in Switzerland

Abstract

SUMMARY The difficulty and the high cost to assess the subsurface properties led to the development of several geophysical techniques. Generally, the focus of a site study is the reconstruction of the S-wave velocity profile down to few tens to hundreds of metres (e.g. 30–300 m), but not the investigation of deeper structures, such as the transition to the crystalline basement. However, such deeper structures are of interest when seismic hazard products have to relate to a reference rock-velocity profile, for example in regional seismic hazard assessment and microzonation studies. To estimate the S-wave velocity profiles down to several kilometres, we study the potential of Rayleigh and Love waves at low (down to 0.1 Hz) and high (up to 20 Hz) frequencies using two seismic arrays of increasing size. The small array, with a maximum inter-station distance of 900 m and a recording time of 3 hr, was aimed at constraining the shallow subsurface down to about 350–400 m, while the big one, with a maximum inter-station distance of more than 29 km and 23 hr of recording had the goal to constrain the deeper structure. The arrays were deployed in northern Switzerland (east of the village of Herdern) within the Swiss Molasse basin, a sedimentary basin north of the Alps stretching from Lake Constance to Lake Geneva; its thickness increases from 800 to 900 m in the northeast to more than 5 km in the southwest. The seismic data recorded by the two arrays were analysed using the techniques developed for the analysis of small-aperture arrays. The results were inverted for the S-wave velocity profile in two steps: first, the Rayleigh and Love wave phase dispersion curves were inverted together. Secondly, the previous dispersion curves were jointly inverted with the measured Rayleigh wave ellipticity angle. The resulting S-wave velocity profiles are similar and show agreement with the available geological and geophysical data, confirming the potential of surface waves to investigate deep structures. Moreover, our analysis proves the feasibility of site characterization techniques to large arrays and the possibility to estimate the P- and S-wave velocity profiles down to 5 km, deeper than the contrast between Molasse basin and crystalline rock at around 2.1 km.